Ocean Engineering最新文献_第5页

Offshore wind power prediction based on improved Hankel-DMD-LSTM hybrid model: Strengthening mode selection and residual compensation mechanism 基于改进Hankel-DMD-LSTM混合模型的海上风电预测：强化模式选择和剩余补偿机制

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-11 DOI: 10.1016/j.oceaneng.2026.124607

Tengjiao Guo , Qiang Liu , Xingyu Fang , Mingkun Fang , Yanzhao Wu , Ran Tao , Ruofu Xiao

Offshore wind energy, characterized by resource potential and sustainable attributes, holds strategic importance in the global energy transition. Despite its advantages of high energy density, stable wind regimes, and reduced land-use conflicts, current exploitation levels remain limited. This is due to the difficulty of offshore units and the volatility and intermittency of wind energy, which lead to non-stationary and multi-scale fluctuations in wind power signals, posing significant challenges for accurate forecasting. To address this, we propose strengthening Hankel-DMD (Hankel Dynamic Mode Decomposition) by incorporating an improved discrete-time coefficient weighting (DTW) criterion, which retains physically interpretable modes to enhance analyzability. Furthermore, to compensate for nonlinear dynamics overlooked during mode decomposition, a Long Short-Term Memory (LSTM) network is integrated into the framework, establishing a physics-data hybrid model that synergizes linear dynamics identification with data-driven residual correction to enhance accuracy. We present a hybrid forecasting framework that combines domain knowledge with data-driven models to provide both improved accuracy and physical interpretability. Applied to real offshore wind data, the method attains a post-compensation R² = 0.937 and lowers overall forecasting error below 8%, yielding engineering-grade short-term forecasts suitable for grid operation and planning.

海上风能具有资源潜力和可持续发展的特点，在全球能源转型中具有重要的战略意义。尽管它具有高能量密度、稳定的风力和减少土地使用冲突的优势，但目前的开发水平仍然有限。这是由于海上发电机组的难度和风能的波动性和间歇性，导致风电信号的非平稳和多尺度波动，对准确预测提出了重大挑战。为了解决这个问题，我们建议通过结合改进的离散时间系数加权（DTW）标准来加强Hankel- dmd （Hankel动态模式分解），该标准保留了物理可解释模式以增强可分析性。此外，为了补偿模式分解过程中忽略的非线性动力学，将长短期记忆（LSTM）网络集成到框架中，建立了物理-数据混合模型，该模型将线性动力学识别与数据驱动的残差校正协同起来，以提高精度。我们提出了一个混合预测框架，将领域知识与数据驱动模型相结合，以提供更高的准确性和物理可解释性。将该方法应用于实际海上风电数据，补偿后的预测R2 = 0.937，总体预测误差低于8%，可获得工程级短期预测，适合电网运行和规划。

{"title":"Offshore wind power prediction based on improved Hankel-DMD-LSTM hybrid model: Strengthening mode selection and residual compensation mechanism","authors":"Tengjiao Guo , Qiang Liu , Xingyu Fang , Mingkun Fang , Yanzhao Wu , Ran Tao , Ruofu Xiao","doi":"10.1016/j.oceaneng.2026.124607","DOIUrl":"10.1016/j.oceaneng.2026.124607","url":null,"abstract":"<div><div>Offshore wind energy, characterized by resource potential and sustainable attributes, holds strategic importance in the global energy transition. Despite its advantages of high energy density, stable wind regimes, and reduced land-use conflicts, current exploitation levels remain limited. This is due to the difficulty of offshore units and the volatility and intermittency of wind energy, which lead to non-stationary and multi-scale fluctuations in wind power signals, posing significant challenges for accurate forecasting. To address this, we propose strengthening Hankel-DMD (Hankel Dynamic Mode Decomposition) by incorporating an improved discrete-time coefficient weighting (DTW) criterion, which retains physically interpretable modes to enhance analyzability. Furthermore, to compensate for nonlinear dynamics overlooked during mode decomposition, a Long Short-Term Memory (LSTM) network is integrated into the framework, establishing a physics-data hybrid model that synergizes linear dynamics identification with data-driven residual correction to enhance accuracy. We present a hybrid forecasting framework that combines domain knowledge with data-driven models to provide both improved accuracy and physical interpretability. Applied to real offshore wind data, the method attains a post-compensation R<sup>2</sup> = 0.937 and lowers overall forecasting error below 8%, yielding engineering-grade short-term forecasts suitable for grid operation and planning.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"352 ","pages":"Article 124607"},"PeriodicalIF":5.5,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192091","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A super-twisting non-singular terminal sliding mode control strategy based on particle swarm optimization for underwater two-link manipulator trajectory tracking 基于粒子群优化的超扭转非奇异末端滑模水下双连杆机械臂轨迹跟踪控制策略

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-11 DOI: 10.1016/j.oceaneng.2026.124581

Xingyao Wang , Yanjun Liu , Gang Xue , Fagang Bai , Shuting Huang

To enhance trajectory tracking accuracy and eliminate control torque chattering in underwater manipulators, this study proposes a novel particle swarm optimization-based super-twisting non-singular terminal sliding mode control (PSO-STNTSMC) scheme. It integrates the finite-time convergence of non-singular terminal sliding mode control and the chattering suppression capability of the super-twisting algorithm. Particle swarm optimization (PSO) is incorporated to tune controller parameters for optimal control performance. Lyapunov stability analysis derives super-twisting gain boundary conditions and establishes theoretical guarantees. The PSO-STNTSMC demonstrates superior tracking performance with significantly diminished error and substantial suppression of chattering phenomena. Compared with PSO-based sliding mode control (PSO-SMC) and PSO-based non-singular terminal sliding mode control (PSO-NTSMC), the PSO-STNTSMC has marked reductions in Root Mean Squared Error (RMSE), ranging from 43.14% to 95.53% across both joints. To further assess the control capability of the proposed methodology, trajectory tracking experiments were performed on a two-joint manipulator in still water. Results demonstrate PSO-STNTSMC outperforms both PSO-SMC and PSO-NTSMC in tracking performance, exhibiting significantly reduced tracking errors, reducing both joints' RMSE by 33.85-69.29%. Compared with the unoptimized STNTSMC in this study, PSO-STNTSMC also achieves RMSE reductions of 27.01% and 27.12% for both joints. These outcomes confirm the efficacy and accuracy of the proposed approach.

为了提高水下机械臂的轨迹跟踪精度，消除控制力矩抖振，提出了一种基于粒子群优化的超扭转非奇异终端滑模控制（PSO-STNTSMC）方案。它结合了非奇异终端滑模控制的有限时间收敛性和超扭转算法的抖振抑制能力。采用粒子群算法对控制器参数进行优化，使控制性能达到最优。李雅普诺夫稳定性分析导出了超扭增益边界条件，并建立了理论保证。PSO-STNTSMC具有优异的跟踪性能，显著减小了误差，有效抑制了抖振现象。与基于pso的滑模控制（PSO-SMC）和基于pso的非奇异终端滑模控制（PSO-NTSMC）相比，PSO-STNTSMC显著降低了两个关节的均方根误差（RMSE），范围在43.14% ~ 95.53%之间。为了进一步评估该方法的控制能力，在静水中对一个双关节机械臂进行了轨迹跟踪实验。结果表明，PSO-STNTSMC的跟踪性能优于PSO-SMC和PSO-NTSMC，跟踪误差显著降低，两种关节的RMSE均降低了33.85 ~ 69.29%。与本研究未优化的STNTSMC相比，PSO-STNTSMC对两个关节的RMSE也分别降低了27.01%和27.12%。这些结果证实了所提出方法的有效性和准确性。

{"title":"A super-twisting non-singular terminal sliding mode control strategy based on particle swarm optimization for underwater two-link manipulator trajectory tracking","authors":"Xingyao Wang , Yanjun Liu , Gang Xue , Fagang Bai , Shuting Huang","doi":"10.1016/j.oceaneng.2026.124581","DOIUrl":"10.1016/j.oceaneng.2026.124581","url":null,"abstract":"<div><div>To enhance trajectory tracking accuracy and eliminate control torque chattering in underwater manipulators, this study proposes a novel particle swarm optimization-based super-twisting non-singular terminal sliding mode control (PSO-STNTSMC) scheme. It integrates the finite-time convergence of non-singular terminal sliding mode control and the chattering suppression capability of the super-twisting algorithm. Particle swarm optimization (PSO) is incorporated to tune controller parameters for optimal control performance. Lyapunov stability analysis derives super-twisting gain boundary conditions and establishes theoretical guarantees. The PSO-STNTSMC demonstrates superior tracking performance with significantly diminished error and substantial suppression of chattering phenomena. Compared with PSO-based sliding mode control (PSO-SMC) and PSO-based non-singular terminal sliding mode control (PSO-NTSMC), the PSO-STNTSMC has marked reductions in Root Mean Squared Error (RMSE), ranging from 43.14% to 95.53% across both joints. To further assess the control capability of the proposed methodology, trajectory tracking experiments were performed on a two-joint manipulator in still water. Results demonstrate PSO-STNTSMC outperforms both PSO-SMC and PSO-NTSMC in tracking performance, exhibiting significantly reduced tracking errors, reducing both joints' RMSE by 33.85-69.29%. Compared with the unoptimized STNTSMC in this study, PSO-STNTSMC also achieves RMSE reductions of 27.01% and 27.12% for both joints. These outcomes confirm the efficacy and accuracy of the proposed approach.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"352 ","pages":"Article 124581"},"PeriodicalIF":5.5,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146154158","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

ADE-MIFS: Real-time cooperative navigation for multi-USVs in dynamic maritime environments ADE-MIFS：动态海洋环境下多usv实时协同导航

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-11 DOI: 10.1016/j.oceaneng.2026.124487

Shihong Yin, Zhengrong Xiang

Distributed cooperative navigation of multiple unmanned surface vehicles (USVs) in dynamic maritime environments requires fusing heterogeneous perception and communication information while complying with COLREGs, under strict real-time constraints. Existing approaches often rely on fixed fusion weights or training-heavy learning frameworks, leading to degraded robustness as sensing or communication quality and fleet scale vary. This paper proposes an adaptive differential evolution-based multi-source information fusion strategy (ADE-MIFS), where a unified fusion representation integrates reciprocal velocity obstacles (RVO), COLREGs-compliant interaction constraints, laser range vectors (LRV), and goal-alignment (GA) cues, and the corresponding fusion weights are optimized automatically. An adaptive differential evolution optimizer with Thompson-sampling-based operator selection strategy is developed to balance exploration and exploitation. It improves global search efficiency without policy-network training via environment interactions. Extensive simulations with 5–20 USVs demonstrate that ADE-MIFS improves task success rate over DE-MIFS by 5.56%-24.14% as the scale increases, and outperforms other four advanced baselines (e.g., DADWA, PPO-RVO, PPO-ORCA, and IFDSDA) in both collision avoidance and navigation efficiency, while maintaining an average decision latency below 0.05 s for real-time deployment.

动态海洋环境下多艘无人水面舰艇（usv）的分布式协同导航需要在严格的实时性约束下，融合异构感知和通信信息，同时遵守COLREGs。现有的方法通常依赖于固定的融合权重或训练繁重的学习框架，导致鲁棒性随着感知或通信质量和舰队规模的变化而下降。本文提出了一种基于自适应差分进化的多源信息融合策略（ADE-MIFS），该策略将互反速度障碍（RVO）、COLREGs-compliant交互约束、激光距离矢量（LRV）和目标对准（GA）线索集成在一个统一的融合表示中，并自动优化相应的融合权重。为了平衡勘探和开采，提出了一种基于汤普森采样的自适应差分进化优化器。它提高了全局搜索效率，而不需要通过环境交互进行策略网络训练。5-20 usv的大量模拟表明，随着规模的增加，ADE-MIFS的任务成功率比DE-MIFS提高了5.56%-24.14%，并且在避免碰撞和导航效率方面优于其他四个先进基线（例如DADWA， PPO-RVO， PPO-ORCA和IFDSDA），同时保持了低于0.05 s的平均决策延迟实时部署。

{"title":"ADE-MIFS: Real-time cooperative navigation for multi-USVs in dynamic maritime environments","authors":"Shihong Yin, Zhengrong Xiang","doi":"10.1016/j.oceaneng.2026.124487","DOIUrl":"10.1016/j.oceaneng.2026.124487","url":null,"abstract":"<div><div>Distributed cooperative navigation of multiple unmanned surface vehicles (USVs) in dynamic maritime environments requires fusing heterogeneous perception and communication information while complying with COLREGs, under strict real-time constraints. Existing approaches often rely on fixed fusion weights or training-heavy learning frameworks, leading to degraded robustness as sensing or communication quality and fleet scale vary. This paper proposes an adaptive differential evolution-based multi-source information fusion strategy (ADE-MIFS), where a unified fusion representation integrates reciprocal velocity obstacles (RVO), COLREGs-compliant interaction constraints, laser range vectors (LRV), and goal-alignment (GA) cues, and the corresponding fusion weights are optimized automatically. An adaptive differential evolution optimizer with Thompson-sampling-based operator selection strategy is developed to balance exploration and exploitation. It improves global search efficiency without policy-network training via environment interactions. Extensive simulations with 5–20 USVs demonstrate that ADE-MIFS improves task success rate over DE-MIFS by 5.56%-24.14% as the scale increases, and outperforms other four advanced baselines (e.g., DADWA, PPO-RVO, PPO-ORCA, and IFDSDA) in both collision avoidance and navigation efficiency, while maintaining an average decision latency below 0.05 s for real-time deployment.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"352 ","pages":"Article 124487"},"PeriodicalIF":5.5,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192086","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Harbor resonance under coupled shear flow-wave forcing: Non-hydrostatic modelling and mechanism analysis 剪切流波耦合作用下的港口共振：非流体静力模拟与机理分析

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-11 DOI: 10.1016/j.oceaneng.2026.124513

Dong Shao , Gang Wang , Junyi Chen , Yuqian Huang , Junsheng Wang , Zhan Wang

Harbor resonance under coupled shear flow-wave forcing is examined with a three-dimensional non-hydrostatic model, emphasizing excitation pathways and modal modulation under realistic wave–current conditions. After validation against benchmark experiments for harbor oscillations, shallow-water shear flows, and wave–current interaction, idealized simulations are conducted in a rectangular harbor with (i) steady and noisy shear flows, (ii) shear flow superimposed on regular waves, and (iii) shear flow with random wave groups. Numerical responses are compared with theory of Fabrikant (1995) to assess where the idealized growth-rate theory holds or deviates. Steady shear-flow arrival excites low-order modes via a transient disturbance; peak amplitudes scale approximately with the square of the shear-flow velocity and decay exponentially, while velocity perturbations sustain low-amplitude long-term oscillations. With regular waves, shear-induced entrance vortices shorten the effective harbor length, shift resonant frequencies upward, and reshape the amplification curve, but only weakly change the final steady modal pattern. With random wave groups, shear flow enhances nonlinear transfer to low-frequency long waves, broadens the effective resonance band, and strengthens first-mode dominance. These results identify entrance shear flow as an important, often underestimated driver and modulator of long-period harbor oscillations.

采用三维非流体静力模型研究了剪切流波耦合作用下的港口共振，强调了实际波流条件下的激励途径和模态调制。在对港口振荡、浅水剪切流和波流相互作用的基准实验进行验证后，在矩形港口中进行了(i)稳定和嘈杂剪切流，（ii）剪切流叠加在规则波上，以及（iii）剪切流随机波群的理想模拟。数值响应与Fabrikant（1995）的理论进行了比较，以评估理想增长率理论在哪里成立或偏离。稳态剪切流到达通过瞬态扰动激发低阶模态；峰值振幅与剪切流速度的平方成正比，呈指数衰减，而速度扰动维持低振幅的长期振荡。在规则波中，剪切诱导的入口涡缩短了有效港长，使共振频率向上移动，并重塑了放大曲线，但对最终稳定模态的改变很小。在随机波群中，剪切流增强了向低频长波的非线性传递，拓宽了有效共振带，增强了第一模态优势。这些结果表明，入口剪切流是一个重要的、经常被低估的长周期港口振荡的驱动和调制器。

{"title":"Harbor resonance under coupled shear flow-wave forcing: Non-hydrostatic modelling and mechanism analysis","authors":"Dong Shao , Gang Wang , Junyi Chen , Yuqian Huang , Junsheng Wang , Zhan Wang","doi":"10.1016/j.oceaneng.2026.124513","DOIUrl":"10.1016/j.oceaneng.2026.124513","url":null,"abstract":"<div><div>Harbor resonance under coupled shear flow-wave forcing is examined with a three-dimensional non-hydrostatic model, emphasizing excitation pathways and modal modulation under realistic wave–current conditions. After validation against benchmark experiments for harbor oscillations, shallow-water shear flows, and wave–current interaction, idealized simulations are conducted in a rectangular harbor with (i) steady and noisy shear flows, (ii) shear flow superimposed on regular waves, and (iii) shear flow with random wave groups. Numerical responses are compared with theory of Fabrikant (1995) to assess where the idealized growth-rate theory holds or deviates. Steady shear-flow arrival excites low-order modes via a transient disturbance; peak amplitudes scale approximately with the square of the shear-flow velocity and decay exponentially, while velocity perturbations sustain low-amplitude long-term oscillations. With regular waves, shear-induced entrance vortices shorten the effective harbor length, shift resonant frequencies upward, and reshape the amplification curve, but only weakly change the final steady modal pattern. With random wave groups, shear flow enhances nonlinear transfer to low-frequency long waves, broadens the effective resonance band, and strengthens first-mode dominance. These results identify entrance shear flow as an important, often underestimated driver and modulator of long-period harbor oscillations.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"352 ","pages":"Article 124513"},"PeriodicalIF":5.5,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192087","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Formation mechanism and earth pressure quantification of internal soil arching during caisson penetration 沉箱侵彻过程中内土拱形成机理及土压力量化

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-11 DOI: 10.1016/j.oceaneng.2026.124521

Denghui Ma , Zhengyin Cai , Kaifang Fan , Shuo Wang

Accurate evaluation of penetration resistance is essential to avoid installation failure for offshore wind turbine caisson foundations. During penetration, the soil arching effect significantly increases the internal earth pressure and shaft friction. This study investigates the arching mechanism and its impact on earth pressure through experimental penetration tests and numerical simulations. Results show that the soil adjacent to the caisson wall moves obliquely downward and away from the wall during penetration. The horizontal displacement of the internal soil plug undergoes a process of an initial increase, subsequent reduction, and final stabilization. As penetration proceeds, principal stresses increase and the major principal stress rotates toward the horizontal, forming a downward-convex arch of major principal stress within the caisson. This arching mechanism amplifies the earth pressure acting on the wall. Based on the identified internal soil arching mechanism, a calculation method for internal earth pressure that explicitly incorporates the soil arching effect is proposed. The results from this method demonstrate excellent consistency with numerical simulations. Furthermore, a criterion is established for determining the critical depth at which internal soil arching initiates, which effectively describes the influences of caisson diameter, internal friction angle, and soil element rotation angle.

准确评估海上风电沉箱基础的抗侵彻能力是避免海上风电沉箱基础安装失效的关键。在侵彻过程中，土拱效应显著增加了土体内部压力和轴摩擦。通过侵彻试验和数值模拟，研究了拱起机理及其对土压力的影响。结果表明：在沉井过程中，靠近井壁的土体呈倾斜向下移动，远离井壁。内部土塞的水平位移经历了一个先增大后减小，最后趋于稳定的过程。随着侵彻的进行，主应力增大，主应力向水平方向旋转，在沉箱内形成一个向下的主应力凸拱。这种拱起机制放大了作用在墙体上的土压力。在确定土拱内部机理的基础上，提出了明确考虑土拱效应的内土压力计算方法。该方法的计算结果与数值模拟结果具有良好的一致性。建立了确定土内拱临界深度的判据，有效地描述了沉箱直径、内摩擦角和土元旋转角的影响。

{"title":"Formation mechanism and earth pressure quantification of internal soil arching during caisson penetration","authors":"Denghui Ma , Zhengyin Cai , Kaifang Fan , Shuo Wang","doi":"10.1016/j.oceaneng.2026.124521","DOIUrl":"10.1016/j.oceaneng.2026.124521","url":null,"abstract":"<div><div>Accurate evaluation of penetration resistance is essential to avoid installation failure for offshore wind turbine caisson foundations. During penetration, the soil arching effect significantly increases the internal earth pressure and shaft friction. This study investigates the arching mechanism and its impact on earth pressure through experimental penetration tests and numerical simulations. Results show that the soil adjacent to the caisson wall moves obliquely downward and away from the wall during penetration. The horizontal displacement of the internal soil plug undergoes a process of an initial increase, subsequent reduction, and final stabilization. As penetration proceeds, principal stresses increase and the major principal stress rotates toward the horizontal, forming a downward-convex arch of major principal stress within the caisson. This arching mechanism amplifies the earth pressure acting on the wall. Based on the identified internal soil arching mechanism, a calculation method for internal earth pressure that explicitly incorporates the soil arching effect is proposed. The results from this method demonstrate excellent consistency with numerical simulations. Furthermore, a criterion is established for determining the critical depth at which internal soil arching initiates, which effectively describes the influences of caisson diameter, internal friction angle, and soil element rotation angle.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"352 ","pages":"Article 124521"},"PeriodicalIF":5.5,"publicationDate":"2026-02-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146192088","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Adaptive prescribed-time stabilization of uncertain nonlinear time-varying systems and its application in ship steering system 不确定非线性时变系统的自适应定时镇定及其在船舶操舵系统中的应用

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-11 DOI: 10.1016/j.oceaneng.2026.124545

Kai Cui, You Wu

For a class of nonlinear time-varying systems with unknown control directions, this paper overcomes two challenging limitations in its adaptive prescribed-time control: the singularity caused by infinite control magnitude and the poor steady-state performance beyond the prescribed time. By introducing an improved time-varying gain function, a coordinate transformation and type-B Nussbaum function, a singularity-free controller is designed. It is proved that system state converges to zero before the prescribed time and maintains zero after that. Two simulation examples including a Norrbin ship steering system are provided to demonstrate the effectiveness and advantages of the control scheme.

针对一类控制方向未知的非线性时变系统，克服了其自适应规定时间控制的两个具有挑战性的局限性：控制量无穷大引起的奇异性和超出规定时间后稳态性能较差。通过引入改进的时变增益函数、坐标变换和b型Nussbaum函数，设计了无奇异控制器。证明了系统状态在规定时间前收敛于零，在规定时间后保持零。以Norrbin船舶转向系统为例进行了仿真，验证了该控制方案的有效性和优越性。

引用次数: 0

Experimental and numerical study on the four-point bending behaviour of small-scale steel pipes with dents 带凹痕小钢管四点弯曲特性的试验与数值研究

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-10 DOI: 10.1016/j.oceaneng.2026.124500

Utkarsh Bhardwaj , Baiqiao Chen, C. Guedes Soares

This paper investigates the structural behaviour of small-diameter steel pipes with dents subjected to four-point bending. Experiments are conducted to determine the influence of small pre-induced dents on the pipes’ bending performance and failure characteristics. The tests involve measuring the load-displacement response and identifying the critical failure modes. Numerical simulations using the finite element method are performed to model the bending process and validate the experimental results. The simulations provide insights into the stress distribution and deformation patterns of the dented pipes. The comparison between experimental and numerical results demonstrates a good correlation, indicating the accuracy of the simulation model. This research contributes to understanding the effects of dents on the structural integrity of steel pipes and offers a validated approach for future predictive modelling.

本文研究了带凹痕的小直径钢管在四点弯曲作用下的结构性能。通过试验研究了预诱导小凹痕对管材弯曲性能和破坏特性的影响。试验包括测量荷载-位移响应和识别临界破坏模式。利用有限元方法对弯曲过程进行了数值模拟，验证了实验结果。模拟提供了对凹痕管道的应力分布和变形模式的深入了解。实验结果与数值结果的对比表明，仿真模型具有良好的相关性，表明了仿真模型的准确性。这项研究有助于理解凹痕对钢管结构完整性的影响，并为未来的预测建模提供了一种有效的方法。

引用次数: 0

Numerical simulation-based collision risk assessment of autonomous underwater vehicles during launch and recovery 基于数值模拟的自主水下航行器发射与回收碰撞风险评估

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-10 DOI: 10.1016/j.oceaneng.2026.124577

Ji Li , Yang Liu , Zhuang Kang , Mingxin Li , Jichuan Kang

This paper presents a model for assessing collision risk during the launch and recovery (L&R) of Autonomous Underwater Vehicles (AUVs), aiming to provide auxiliary information to operators. The study focuses on a specific type of AUV side crane Launch and Recovery System (LARS), investigating the nonlinear coupling interactions among the ship, cable, and AUV motion under irregular wave conditions. A dynamic model of LARS is developed, considering factors such as seawater resistance and cable length variations. The motion of the AUV under different sea conditions is calculated and analyzed. A collision-probability estimation model for L&R is formulated from extrapolated AUV dynamic responses. Based on the Bernoulli trial principle, the collision probability is calculated through batch simulation, and the half-width of the confidence interval and the chi-square statistic are computed to assess the accuracy and stability of the results. Finally, the collision probabilities of different dynamic models of the LARS are analyzed and compared, and a risk assessment is conducted in accordance with DNVGL-RP-N103. The results provide a basis for evaluating the feasibility of AUV L&R operations under different sea conditions. This basis serves as auxiliary information for LARS operators of AUV, helping to make decisions during operation.

本文提出了一种自主水下航行器（auv）发射与回收（L&；R）碰撞风险评估模型，旨在为操作者提供辅助信息。研究了一种特殊类型的水下航行器侧吊发射和回收系统（LARS），研究了不规则波浪条件下船舶、电缆和水下航行器运动之间的非线性耦合相互作用。考虑海水阻力和电缆长度变化等因素，建立了LARS的动态模型。对水下航行器在不同海况下的运动进行了计算和分析。根据外推的AUV动态响应，建立了L&；R的碰撞概率估计模型。基于伯努利试验原理，通过批量模拟计算碰撞概率，并计算置信区间的半宽度和卡方统计量，以评估结果的准确性和稳定性。最后，对不同动态模型下LARS的碰撞概率进行了分析比较，并按照DNVGL-RP-N103进行了风险评估。研究结果为评价不同海况下AUV起降作业的可行性提供了依据。该基础可作为AUV的LARS操作员的辅助信息，有助于在操作过程中做出决策。

{"title":"Numerical simulation-based collision risk assessment of autonomous underwater vehicles during launch and recovery","authors":"Ji Li , Yang Liu , Zhuang Kang , Mingxin Li , Jichuan Kang","doi":"10.1016/j.oceaneng.2026.124577","DOIUrl":"10.1016/j.oceaneng.2026.124577","url":null,"abstract":"<div><div>This paper presents a model for assessing collision risk during the launch and recovery (L&R) of Autonomous Underwater Vehicles (AUVs), aiming to provide auxiliary information to operators. The study focuses on a specific type of AUV side crane Launch and Recovery System (LARS), investigating the nonlinear coupling interactions among the ship, cable, and AUV motion under irregular wave conditions. A dynamic model of LARS is developed, considering factors such as seawater resistance and cable length variations. The motion of the AUV under different sea conditions is calculated and analyzed. A collision-probability estimation model for L&R is formulated from extrapolated AUV dynamic responses. Based on the Bernoulli trial principle, the collision probability is calculated through batch simulation, and the half-width of the confidence interval and the chi-square statistic are computed to assess the accuracy and stability of the results. Finally, the collision probabilities of different dynamic models of the LARS are analyzed and compared, and a risk assessment is conducted in accordance with DNVGL-RP-N103. The results provide a basis for evaluating the feasibility of AUV L&R operations under different sea conditions. This basis serves as auxiliary information for LARS operators of AUV, helping to make decisions during operation.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"352 ","pages":"Article 124577"},"PeriodicalIF":5.5,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191993","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Dynamic obstacle avoidance algorithm for UUV based on physics prior LSTM network model and dynamic window approach 基于物理先验LSTM网络模型和动态窗口法的UUV动态避障算法

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-10 DOI: 10.1016/j.oceaneng.2026.124503

Haomiao Yu, Zeyuan Zhang

Unmanned underwater vehicle (UUV) is increasingly required to operate safely in dynamic and uncertain marine environment, where limited sensing range and complex obstacle motions make real-time collision avoidance particularly challenging. To address these issues, a dynamic obstacle avoidance algorithm that integrates a physics prior LSTM network model (PPLNM) and dynamic window approach (DWA) is proposed in this paper to enhance the safety and efficiency of the operation. The PPLNM combines a constant-acceleration physical model with an LSTM network to improve the accuracy of multi-step trajectory prediction and to ensure the physical interpretability of the predicted motions. To explicitly consider prediction uncertainty, future obstacle positions are further described using spherical uncertainty regions under a prescribed confidence level. These spherical regions are then incorporated into an improved DWA algorithm, thereby realizing an anticipatory dynamic obstacle avoidance algorithm. To evaluate the effectiveness of the proposed algorithm, simulation experiments are conducted under various dynamic scenarios. The results show that the PPLNM achieves significantly lower prediction errors than both the Kalman Filter-based predictor and the LSTM-only predictor, and that the PPLNM-enhanced DWA generates smoother and safer trajectories than the standard DWA.

无人水下航行器（UUV）越来越需要在动态和不确定的海洋环境中安全运行，在这些环境中，有限的传感范围和复杂的障碍物运动使得实时防撞变得尤其具有挑战性。为了解决这些问题，本文提出了一种将物理先验LSTM网络模型（PPLNM）与动态窗口方法（DWA）相结合的动态避障算法，以提高避障操作的安全性和效率。PPLNM将恒加速度物理模型与LSTM网络相结合，提高了多步轨迹预测的精度，并保证了预测运动的物理可解释性。为了明确考虑预测的不确定性，在规定的置信水平下，进一步使用球面不确定性区域描述未来障碍物位置。然后将这些球面区域纳入改进的DWA算法，从而实现预期的动态避障算法。为了评估该算法的有效性，在各种动态场景下进行了仿真实验。结果表明，与基于卡尔曼滤波的预测器和仅基于lstm的预测器相比，PPLNM的预测误差显著降低；与标准DWA相比，PPLNM增强的DWA生成的轨迹更平滑、更安全。

{"title":"Dynamic obstacle avoidance algorithm for UUV based on physics prior LSTM network model and dynamic window approach","authors":"Haomiao Yu, Zeyuan Zhang","doi":"10.1016/j.oceaneng.2026.124503","DOIUrl":"10.1016/j.oceaneng.2026.124503","url":null,"abstract":"<div><div>Unmanned underwater vehicle (UUV) is increasingly required to operate safely in dynamic and uncertain marine environment, where limited sensing range and complex obstacle motions make real-time collision avoidance particularly challenging. To address these issues, a dynamic obstacle avoidance algorithm that integrates a physics prior LSTM network model (PPLNM) and dynamic window approach (DWA) is proposed in this paper to enhance the safety and efficiency of the operation. The PPLNM combines a constant-acceleration physical model with an LSTM network to improve the accuracy of multi-step trajectory prediction and to ensure the physical interpretability of the predicted motions. To explicitly consider prediction uncertainty, future obstacle positions are further described using spherical uncertainty regions under a prescribed confidence level. These spherical regions are then incorporated into an improved DWA algorithm, thereby realizing an anticipatory dynamic obstacle avoidance algorithm. To evaluate the effectiveness of the proposed algorithm, simulation experiments are conducted under various dynamic scenarios. The results show that the PPLNM achieves significantly lower prediction errors than both the Kalman Filter-based predictor and the LSTM-only predictor, and that the PPLNM-enhanced DWA generates smoother and safer trajectories than the standard DWA.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"352 ","pages":"Article 124503"},"PeriodicalIF":5.5,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191994","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Evaluating the impact force of submarine landslide on pipelines: insights from flume tests and multiphase flow model analysis 海底滑坡对管道的冲击力评估：水槽试验和多相流模型分析的启示

IF 5.5 2区工程技术 Q1 ENGINEERING, CIVIL

Ocean Engineering

Pub Date : 2026-02-10 DOI: 10.1016/j.oceaneng.2026.124629

Zili Dai, Chong Wang, Renfeng Peng, Mingtao Jiang

Submarine landslides, characterized by vast volumes, high velocities, and extensive coverage, pose significant threats to marine pipelines. This study systematically assesses impact loads exerted by submarine landslides on pipelines via flume model tests and multiphase flow numerical analysis. The results reveal that impact loads consist of a drag force component along the landslide movement direction and a lift force component perpendicular to the sliding direction. Under identical conditions, the peak lift force is approximately 50-70% of the peak drag force. Both peak drag and lift forces exhibit a linear relationship with the square of landslide velocity, though the drag peak occurs significantly earlier. Additionally, key parameters such as landslide mass volume, slope gradient, pipeline diameter, and span height exert a significant influence on the peak load magnitudes. The quantitative impact load evaluation model established in this study demonstrates strong consistency between numerical predictions and experimental data, with correlation coefficients ranging from 0.84 to 0.98. These findings provide a critical theoretical basis and technical support for disaster risk assessment, as well as the optimized design of pipeline protective engineering measures.

海底滑坡具有体积大、速度快、覆盖范围广等特点，对海洋管道构成重大威胁。通过水槽模型试验和多相流数值分析，系统地评估了海底滑坡对管道的冲击载荷。结果表明，冲击荷载由沿滑坡运动方向的阻力分量和垂直于滑坡运动方向的升力分量组成。在相同条件下，峰值升力约为峰值阻力的50-70%。阻力峰值和升力峰值均与滑坡速度的平方成线性关系，但阻力峰值出现的时间明显提前。此外，滑坡体体积、坡度、管径、跨高等关键参数对峰值荷载大小有显著影响。本文建立的冲击载荷定量评价模型与实验数据具有较强的一致性，相关系数在0.84 ~ 0.98之间。研究结果为灾害风险评估以及管道防护工程措施的优化设计提供了重要的理论依据和技术支持。

{"title":"Evaluating the impact force of submarine landslide on pipelines: insights from flume tests and multiphase flow model analysis","authors":"Zili Dai, Chong Wang, Renfeng Peng, Mingtao Jiang","doi":"10.1016/j.oceaneng.2026.124629","DOIUrl":"10.1016/j.oceaneng.2026.124629","url":null,"abstract":"<div><div>Submarine landslides, characterized by vast volumes, high velocities, and extensive coverage, pose significant threats to marine pipelines. This study systematically assesses impact loads exerted by submarine landslides on pipelines via flume model tests and multiphase flow numerical analysis. The results reveal that impact loads consist of a drag force component along the landslide movement direction and a lift force component perpendicular to the sliding direction. Under identical conditions, the peak lift force is approximately 50-70% of the peak drag force. Both peak drag and lift forces exhibit a linear relationship with the square of landslide velocity, though the drag peak occurs significantly earlier. Additionally, key parameters such as landslide mass volume, slope gradient, pipeline diameter, and span height exert a significant influence on the peak load magnitudes. The quantitative impact load evaluation model established in this study demonstrates strong consistency between numerical predictions and experimental data, with correlation coefficients ranging from 0.84 to 0.98. These findings provide a critical theoretical basis and technical support for disaster risk assessment, as well as the optimized design of pipeline protective engineering measures.</div></div>","PeriodicalId":19403,"journal":{"name":"Ocean Engineering","volume":"352 ","pages":"Article 124629"},"PeriodicalIF":5.5,"publicationDate":"2026-02-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0